Novel frameshift mutations of ANKUB1, GLI3, and TAS2R3 associated with polysyndactyly in a Chinese family

Abstract Background Polysyndactyly (PSD) is an autosomal dominant genetic limb malformation caused by mutations. Methods Whole exome sequencing and Sanger sequencing were used to determine the mutations in PSD patients. Luciferase reporter assay was performed to determine the effect of GLI3 mutation on its transcriptional activity. Results In this study, we investigated the gene mutations of three affected individuals across three generations. The frameshift mutations of GLI3 (NM_000168:c.4659del, NP_000159.3: p.Ser1553del), ANKUB1 (NM_001144960:c.1385del, NP_001138432.1: p.Pro462del), and TAS2R3 (NM_016943:c.128_131del, NP_058639.1: p.Leu43del) were identified in the three affected individuals, but not in three unaffected members by whole exome sequencing and sanger sequencing. Luciferase reporter assay demonstrated that GLI3 mutation reduced the transcriptional activity of GLI3. The results from SMART analysis showed that the frameshift mutation of TAS2R3 altered most protein sequence, which probably destroyed protein function. Although the frameshift mutation of ANKUB1 did not locate in ankyrin repeat domain and ubiquitin domain, it might influence the interaction between ANKUB1 and other proteins, and further affected the ubiquitinylation. Conclusion These results indicated that the frameshift mutations of GLI3, ANKUB1, and TAS2R3 might alter the functions of these proteins, and accelerated PSD progression.


| Patients and ethical compliance
A three-generation Chinese family that contains three patients with limb malformation and three healthy individuals was investigated. Venous blood samples were obtained from affected and unaffected family members. Genomic DNA was isolated and quantified using Qubit® DNA Assay Kit in Qubit® 2.0 Flurometer (Life Technologies). This study was approved by the Ethics Committee of Shandong provincial Hospital affiliated to Shandong University, and informed consent was obtained for experimentation with human subjects.

| Library preparation and sequencing
Genomic DNA was sheared into 150-200 bp fragments by hydrodynamic shearing system (Covaris). Remaining overhangs were converted into blunt ends using exonuclease/ polymerase activities. After adenylation, DNA fragments were ligated with adapter oligonucleotides on both ends, and were selectively enriched via a PCR reaction, followed by hybridization with biotin-labeled probe. Magnetic beads with streptomycin were used to capture the exons of genes. DNA libraries were enriched in a PCR reaction and sequenced on Illumina Hiseq platform.

| Variant analysis
Valid sequencing data was mapped to the human reference genome sequence from UCSC database by Burrows-Wheeler Aligner (BWA) software (Li & Durbin, 2009-1). Singlenucleotide polymorphisms (SNPs) and insertions-deletions (InDels) were identified by Samtools and GATK software (Li et al., 2009-2). 1,000 Genomes databases and dbSNP databases were used to characterize the detected variants. ANNOVAR (Wang, Li, & Hakonarson, 2010) was performed to annotate SNPs and InDels. Gene transcript annotation databases, including Consensus CDS, RefSeq, Ensembl, and UCSC, were used to determine amino acid alternation. Sorting Intolerant from Tolerant (SIFT) and Polymorphism Phenotyping version 2 (PolyPhen-2) were performed to assess the functional relevance of the detected variants.

| Sanger sequencing
The selected mutations were verified by PCR combined with Sanger sequencing. The primers for amplifying DNA fragments containing mutation sites are listed in Table 1.

| Construction of GLI3 truncation mutation
Full-length human GLI3 cDNA was obtained by PCR amplification using the specific primers (CTAGCGGCCGCGCCACCATGGAGGCCCAGTCC; GCTCTAGA TCCTATTGATTTCCGTTGG). According to the results from variant analysis, GLI3 truncation mutation was generated by PCR amplification using the specific primers (CTAGCGGCCGCGCCACCATGGAGGCCCAGTCC; GCTCTAGA TCATGTCCCCGATAGCC). The fragments were digested by Not I and Xba I, and were cloned into flagtagged pcDNA3 vector (Addgene) to construct pcDNA3.1-GLI3-WT (wild type) and pcDNA3.1-GLI3-MT (mutation type).
(Promega) was used for a normalizing control. After 48 hr of incubation, luciferase activities were determined using the Dual-Luciferase Assay (Promega) according to the manufacturer's instructions.

| Western blot analysis
The transfected cells were lysed using RIPA lysis buffer (Thermo) to obtain total protein. The protein concentration was measured using the bicinchoninic acid method. Equal amounts of proteins were loaded on 8% SDS-PAGE and transferred to PVDF membranes (EMD Millipore). The membranes were blocked for 1 hr with 5% nonfat milk, and were then incubated with the indicated primary and secondary antibodies. The protein signals were visualized using the enhanced chemiluminescence method and quantified using Scion Image 4.03 software.

| Statistical analysis
Statistical analysis was carried out using GraphPad Prism 5.0. All data are presented as mean ± SD from at least three independent experiments. The Student's t test was used to assess the difference between two groups. One-way ANOVA was used to assess the difference among multiple groups. p < .05 was considered as statistical significance.

| Clinical features
The pedigree of the Chinese family with PSD was shown in Figure 1a, was close of preaxial polydactyly type IV. In this family, this phenotype affected three successive generations. The proband was a 1.5-year-old boy with PSD at both hands and foots since he was born ( Figure 1b). A simple and incomplete syndactyly was noted at the middle, ring, and little fingers in his left hand. Ulnar polydactyly of little finger (only finger bud) was also observed. Radial polydactyly of thumb showed only an incomplete division at the distal phalanx, and finger appearance was normal. There was syndactyly at the toes 1, 2, 3, and 4 and polydactyly of big toe in his left foot. Syndactyly at the toes 1, 2, and 3 was also found in his right foot. The second affected member was the mother of proband. A simple and incomplete syndactyly was noted at the ring and little fingers in her both hands. Duplicated hallux and syndactyly between toes 2 and 3 were observed in both feet. The third affected member was the grandma of proband. Her hands had no visible abnormalities and her feet appeared the same phenotype as the second affected member (no pictures were captured). Apart from PSD, no other abnormalities were found. The clinical features of the three affected subjects were presented in Table 2.
T A B L E 1 Sequences of the primers used for gene mutations amplification and sequencing

Gene name
Forward primer Reverse primer F I G U R E 1 (a) the pedigree of a three-generation Chinese family with PSD. Arrow represented proband. Circles and squares represented female and male, respectively. Blank and black represented unaffected and affected member, respectively. (b) The clinical characteristics of the proband (III 1) and his mother (II 2)

| Mutation analysis
Whole exome sequencing was performed to detect DNA samples from affected and unaffected family members. The raw image data were obtained and transformed to sequenced reads that was named as raw data. After removing unqualified reads, such as adapter contamination, low-quality nucleotides, and unrecognizable nucleotide, clean data were obtained and mapped to the reference human genome (UCSC, hg19) to generate BAM files by BWA software. SNPs and InDels were identified by Samtools and GATK software, and were annotated by ANNOVAR. According to priority score and quality score for mutations, 147 SNPs and 15 InDels remained.
Considering the important effects of InDels on gene function, we selected 3 InDels with higher quality score, including GLI3 (NM_000168:c.4659del, NP_000159.3: p.Ser1553del), ANKUB1 (NM_001144960:c.1385del, NP_00113 8432.1: p.Pro462del), and TAS2R3 (NM_016943:c.128_131del, NP_058639.1: p.Leu43del). All of these three mutations were heterozygous, and the allele frequencies were listed in Table 3. Direct DNA sequencing showed that mutations in ANKUB1, GLI3, and TAS2R3 were observed in all three affected members, but not in all three unaffected members ( Figure 2), suggesting that these mutations might be associated with PSD. Other SNPs identified in this study were listed in supplemental file 1.

| GLI3 truncation mutation reduced the transcriptional activity of GLI3
Figure 3a illustrated that GLI3 protein is comprised of three parts, among which the N-terminal part contains the zinc finger domain (ZFD, AA, 462-645), the middle part contains the protein cleavage site (PC, AA, 703-740), and the C-terminal part contains two transactivating domains (TA2, AA, 1044-1322 and TA1, AA, 1376-1580). Mutations in N-terminal part cause GCPS phenotype, mutations in middle part lead to PHS, and mutations in C-terminal part result in a loss of activator function in a grade manner, inducing GCPS and polydactyly. Considering that GLI3 is a key transcription factor regulating limb development, we explored whether GLI3 mutation (exon15:c.4659delC, p.S1553fs) affected the transcriptional activity of GLI3. As shown in Figure 3b, no significant difference was observed in molecular weight of GLI3 in 293T cells overexpressing flag-tagged pcDNA3.1-GLI3-WT and flag-tagged pcDNA3.1-GLI3-MT. However, overexpression of GLI3-MT remarkably reduced the luciferase activity mediated by the promoter of PTCH1 compared to overexpression of GLI3-WT (Figure 3c). These results suggested that GLI3 mutation (exon15:c.4659delC, p.S1553fs) reduced the transcriptional activity of GLI3.

| The location of the novel frameshift mutations of ANKUB1 and TAS2R3
We further mapped the putative structural domain according to the protein sequence using SMART analysis (http://smart. embl-heide lberg.de/). Figure 4a illustrated that ANKUB1 was comprised of three ankyrin repeat domains (ANK), one ubiquitin domain (UBQ), and one low complexity (LC). The mutation of ANKUB1 located in front of the LC domain, which resulted in the deficiency of this domain. It was necessary to validate the role of this ANKUB1 mutation in the function of ANKUB1. Figure 4b showed that TAS2R3 contains multiple transmembrane regions. The frameshift mutation of TAS2R3 located in region prior to the second transmembrane domain, which deleted most amino acid sequence of TAS2R3 (after 43rd amino acid). Therefore, it was likely that this mutation destroyed the protein function.

| DISCUSSION
PSD is most common limb deformity and is regulated by multiple genes (Goodman, 2002). HOXD13 mutations have widely been implicated in PSD (Brison et al., 2014).
Recently, GLI3 mutations are also identified in human congenital malformation (Al-Qattan et al., 2017). GLI3 is one of glioma-associated oncogene family members, and acts as a transcription factor that regulates cell proliferation, death, and differentiation (Li, Zhang, Choi, Litingtung, & Chiang, 2004). Therefore, mutations of GLI3 gene cause several adverse developmental consequences. GLI3 mutations have been demonstrated to be closely associated with GCPS, PHS, and isolated polydactyly (Al-Qattan, 2012;Al-Qattan et al., 2017;Demurger et al., 2015). The first mutation of GLI3 has been found in 1991 (Vortkamp, Gessler, & Grzeschik, 1991). Since then, a total of 223 mutations are identified in many kinds of human genetic diseases (Stenson et al., 2017). GLI3 protein can be functionally divided into the N-terminal part, the middle part, and the C-terminal part. The N-terminal part contains the zinc finger domain (ZFD, AA, 462-645), the middle part contains the protein cleavage site (PC, AA, 703-740), and the C-terminal part contains two transactivating domains (TA2, AA, 1044-1322 and TA1, AA, 1376-1580) (Demurger et al., 2015). Mutations in N-terminal part mainly lead to GCPS phenotype, mutations in middle part can give rise to PHS, and mutations in C-terminal part can result in GCPS and polydactyly. In the present study, we found a frameshift mutation of GLI3 in a pedigree of the Chinese family with PSD (preaxial polydactyly type IV). This frameshift  (Ni et al., 2019). Therefore, our results suggested that this frameshift mutation of GLI3 might be a main reason for preaxial PD type IV phenotype in this Chinese family. We also identified other gene mutations, such as ANKUB1 and TAS2R3. The frameshift mutations of ANKUB1 and TAS2R3 have never been identified to be associated with SPD. The full name of ANKUB1 is ankyrin repeat and ubiquitin domain containing 1, also known as C3orf16. As the name implies, ANKUB1 contained ankyrin repeat domain and ubiquitin domain. We mapped the putative structural domain according to the protein sequence using SMART analysis (http://smart.embl-heide lberg.de/). As shown in Figure 4a, ANKUB1 was comprised of three ankyrin repeat domains (ANK), one ubiquitin domain (UBQ), and one low complexity (LC). Ankyrin repeat domain is a 33-residue motif that is frequently found in vertebrate proteins, mediating the protein-protein interaction. Ubiquitin contains 76 amino acid residues that are extremely conserved in all eukaryotic cells. Ubiquitin affects the function of other proteins by ubiquitinylation. In this study, the frameshift at codon 1,385 (NM_001144960, exon5:c.1385delG, p.P462fs) truncated C-terminally the ANKUB1 protein, but did not affect ankyrin repeat domain and ubiquintin domain. It is necessary to determine whether this frameshift mutation altered the function of ANKUB1, and affected osteogenesis and osteoclast differentiation. TAS2R3 is a member of type 2 taste receptors (TAS2Rs) that belong to a class of G protein-coupled receptors (Choi et al., 2018). TAS2R3 contains multiple transmembrane regions (Figure 4b), mediating signal transduction on the cellular membrane. Previous studies mainly focused on the function of TAS2Rs in bitterness sensing (Choi et al., 2018). Recently, genetic variation in TAS2R3 has been associated with the risk of papillary thyroid carcinoma and regulates thyroid function (Choi et al., 2018). In addition, the SNP of TAS2R3 has been closely correlated to male infertility (Gentiluomo et al., 2017). In the present study, a frameshift mutation of TAS2R3 (NM_016943, exon1:c.128_131delTGTC, p.L43fs) was found to be associated with PSD. This mutation altered most amino acid sequence of TAS2R3 (after 43rd amino acid), which probably destroyed the protein function. It would be interesting to determine the role of TAS2R3 in limb development.

F I G U R E 4 The location of ANKUB1
and TAS2R3 mutations. (a) Schematic diagram and protein sequence of ANKUB1 domains and location of ANKUB1 mutation identified in this study. (b) Schematic diagram and protein sequence of TAS2R3 domains and location of TAS2R3 mutation identified in this study

| CONCLUSIONS
In summary, we evidenced that a novel frameshift mutation in GLI3 segregates with the malformation in the Chinese family, but novel rare variants in ANKUB1 and TAS2R3 also segregate in this family. Further investigation demonstrated that the frameshift mutation of GLI3 reduced the transcriptional activity of GLI3. According to the results from SMART analysis, the frameshift mutation of TAS2R3 altered most protein sequence, resulting in a damage in protein function. Although the frameshift mutation of ANKUB1 did not affected ankyrin repeat domain and ubiquitin domain, it probably affected the interaction between ANKUB1 and other proteins, and altered the ubiquitinylation. Our results suggested that the frame-shift mutations of GLI3 reduced its transcriptional activity, contributing to malformation progression. Since the pedigree is small, the causative nature of variants in ANKUB1 and TAS2R3 remains unclear.